O R I G I N A L A R T I C L E
Characterization of Gastric Cardia Tumors: Differences
in
Helicobacter pylori
Strains and Genetic Polymorphisms
De´bora Menezes da Costa1•Eliane dos Santos Pereira1•Isabelle Joyce de Lima Silva-Fernandes1•Ma´rcia Vale´ria Pitombeira Ferreira1•Silvia Helena Barem Rabenhorst1
Received: 20 February 2015 / Accepted: 9 April 2015 / Published online: 24 April 2015
ÓSpringer Science+Business Media New York 2015
Abstract
Background Gastric cancer results from a multifactorial process and is one of the most common causes of death worldwide. These tumors can arise in the distal stomach (non-cardia) and in the cardia region, presenting different characteristics and frequency of occurrence worldwide. Aims To search for differences between tumors of dif-ferent locations that could explain the presence of cardia tumors, consideringHelicobacter pyloristrains and genetic polymorphisms.
Materials and Methods DNA was extracted from gastric adenocarcinoma tissue of 127 patients.Helicobacter pylori genes were detected by PCR, and polymorphisms by PCR– RFLP.
Results Most of the tumors were located in non-cardia. The genotype 28152GA ofXRCC1showed an increase in risk of cardia tumors. In analysis performed considering gender, women carrying TNF-308GA genotype showed a decreased risk of non-cardia tumors, while in men the decreased risk of non-cardia tumors was associated with TNF-308GG genotype. Genotypes combinations showed that the SNPs RAD51 135G[C, XRCC3 18067C[T, and
XRCC128152G[A had some combinations more frequent in cardia tumors, with an increased risk. Patients infected by cagE-positive strains presented a positive correlation with non-cardia tumors.
Conclusion The results showed some susceptibility dif-ferences between tumors of different locations. There was an increased risk relationship between three repair enzyme SNPs and cardia tumors, and the G allele of the cytokine gene TNF negatively influenced the development of non-cardia tumors. Helicobacter pylori strains seemed to be different in the cardia region, where they were less virulent than those located in the distal region of the stomach.
Keywords Cardia tumorsHelicobacter pyloriRepair enzymes Interleukins
Introduction
Gastric cancer is one of the most common causes of death worldwide. The prognosis of this disease is poor, par-ticularly in Brazil, with a 5-year survival rate of 20–30 %, since most cases are diagnosed with advanced stage [1]. While Helicobacter pylori infection is a well-established cause of gastric cancer, different environmental and ge-netic factors (e.g., H. pylori virulence genes and single nucleotide polymorphisms (SNPs) in the host’s genes) are involved in different stages of the cancer process [2]. Gastric tumors can arise in the distal stomach (non-cardia) and in the cardia region, and in the last four decades, there has been an increase in cases of cardia cancer, mainly in developed countries with remarkable geographic aggrega-tion [3, 4]. These tumors seem to be different in some characteristics, and they are now recognized as two dif-ferent clinical entities. However, what predisposes the
Electronic supplementary material The online version of this article (doi:10.1007/s10620-015-3666-0) contains supplementary material, which is available to authorized users.
& De´bora Menezes da Costa
[email protected] Silvia Helena Barem Rabenhorst [email protected]
1 Molecular Genetics Laboratory, Department of Pathology
gastric cancer development in a specific region of the stomach is still not clear [4]. Thus, the objective of this study was to search for differences between tumors of different locations that could explain the presence of cardia tumors, considering H. pylori strains and genetic poly-morphisms of certain genes.
Materials and Methods
Subjects
The present study was approved by the Hospital Ethics Committee of the Federal University of Ceara´, Brazil, and all subjects signed an informed consent form before in-clusion. Adenocarcinoma samples included in this study were obtained from patients who had undergone gastrec-tomy at University Hospital Walter Cantı´deo, Santa Casa de Miserico´rdia Hospital, and General Hospital located in Fortaleza (Ceara´—Brazil). The histopathological data were obtained from histopathological records and reviewed by a pathologist of the team.
DNA Extraction
Genomic DNA was extracted from frozen tumor tissue using the cetyltrimethyl ammonium bromide (CTAB) technique, adapted from Foster and Twell [5]. DNA ex-traction was done only in fragments that showed more than 80 % tumor cells. DNA quality was analyzed by 1 % agarose gel electrophoresis, and quantity was determined using the NanoDropÒ3300 fluorospectrometer (Wilming-ton, DE, USA).
Detection ofH. pyloriInfection andvacA Alleles and the Presence ofcagA, cagE, andvirB11 Genes
Helicobacter pyloriinfection was detected by amplification of theureC gene using primers for PCR, as described by Lage et al. [6]. For theH. pylori-positive samples, thevacA alleles andcagA, cagE, andvirB11 genes were identified by PCR according to Lima et al. [7] (Supplemental Table 1).
Genotyping of Host’s Polymorphisms
The genetic polymorphisms of RAD51 135G[C (rs1801320), XRCC3 18067C[T (rs861539), MLH1 -93G[A (rs1800734), MGMT 533A[G (rs2308327), XRCC128152G[A (rs25487),XPD35931A[C (rs13181), IL1B-511C[T (rs16944), IL1RN (VNTR), TNF-308G[A (rs1800629), IL6-174G[C (rs1800795), and IL8-251A[T
(rs4073) were detected by PCR–RFLP, as described in Supplemental Table 2.
Statistical Analyses
The statistical analyses were carried out using the programs SPSSÒ version 15.0 (Chicago, IL, USA), SNPStats, and UNPHASEDÒ version 3.1.7 (London, UK). Statistically significant differences were evaluated by the Chi-square test (v2) or Fisher exact test. The results were considered
statistically significant whenpwas less than 0.05.
Results
Study Population
Of a total of 127 collected gastric tumor samples, 30 (20.62 %) were located in the cardia region, and 97 (76.38 %) in non-cardia. The patients’ median age was 65 years, independent of tumor location. Among the tu-mors located in the cardia region, there were a significantly higher proportion of males (Table 1).
SNP Analysis
A total of 11 SNPs were analyzed, individually and in groups. The genotype and allele distributions and risk analysis are shown in Table2. Considering all cases, only the heterozygote genotype of XRCC1 28152G[A (Arg399Gln) showed an increase in risk of tumor located in the cardia region. This result was confirmed by SNPStats analysis, where this genotype was associated with a
Table 1 Characteristics of gastric cancer patients according to tumor location
Cardia
N(%)
Non-cardia
N(%)
v2 p
Tumor location 30 (29.62) 97 (76.38) – – Gender
Male (65 %) 26 (20.5) 57 (44.9)
Female (35 %) 4 (3.1) 40 (31.5) 7.88 0.005* Age
\60 6 (4.7) 32 (25.2)
C60 24 (18.9) 65 (51.2) 1.84 0.174 Tumor stage (n=124)
Table 2 Genotype and allelic distribution of gastric cardia and non-cardia tumors
Cardia Non-cardia v2 OR p RAD51 (G[C)
GG 23 76.7 % 77 79.4 % – – –
GC 4 13.3 % 13 13.4 % 0.00 1.03 (0.25–3.87) 1.00 CC 3 10.0 % 7 7.2 % 0.25 1.43 (0.27–6.92) 0.697 G allele 50 83.3 % 167 86.1 %
C allele 10 16.7 % 27 13.9 % 0.28 1.24 (0.52–2.90) 0.597
XRCC3 (C[T)
CC 19 63.3 % 51 52.6 % – – –
CT 8 26.7 % 41 42.3 % 1.92 0.52 (0.19–1.43) 0.165 TT 3 10.0 % 5 5.2 % 0.38 1.61 (0.27–8.90) 0.680 C allele 46 76.7 % 143 73.7 % – – – T allele 14 23.3 % 51 26.3 % 0.21 0.85 (0.41–1.76) 0.646
MLH1 (G[C)
GG 15 50.0 % 61 62.9 % – – –
GC 14 46.7 % 27 27.8 % 2.97 2.11 (0.82–5.42) 0.085 CC 1 3.3 % 9 9.3 % 0.55 0.45 (0.02–4.03) 0.680 G allele 44 73.3 % 149 76.8 % – – – C allele 16 26.7 % 45 23.2 % 0.30 1.20 (0.59–2.45) 0.582
MGMT (A[G)
AA 25 83.3 % 74 76.3 % – – –
AG 4 13.3 % 15 15.5 % 0.15 0.79 (0.20–2.88) 1.00 GG 1 3.3 % 8 8.2 % 0.90 0.37 (0.02–3.19) 0.684 A allele 54 90.0 % 163 84.0 % – – – G allele 6 10.0 % 31 16.0 % 1.32 0.58 (0.21–1.57) 0.251
XRCC1 (G[A)
GG 12 40.0 % 53 55.2 % – – –
GA 16 53.3 % 28 29.2 % 4.41 2.52 (0.96–6.66) 0.035* AA 2 6.7 % 15 15.6 % 0.43 0.59 (0.08–3.29) 0.723 G allele 40 66.7 % 134 69.8 % – – – A allele 20 33.3 % 58 30.2 % 0.21 1.16 (0.59–2.24) 0.647
XPD (A[C)
AA 13 43.3 % 53 54.6 % – – –
AC 10 33.3 % 34 35.1 % 0.15 1.20 (0.43–3.34) 0.701 CC 7 23.3 % 10 10.3 % 3.41 2.85 (0.79–10.32) 0.107 A allele 36 60.0 % 140 72.2 % – – – C allele 24 40.0 % 54 27.8 % 3.19 1.73 (0.90–3.30) 0.104
IL1B (-511C[T)
CC 10 34.5 % 26 27.1 % – – –
CT 12 41.4 % 54 56.3 % 1.27 0.58 (0.20–1.68) 0.260 TT 7 24.1 % 16 16.7 % 0.05 1.14 (0.31–4.15) 0.826 C allele 32 55.2 % 106 55.2 % – – – T allele 26 44.8 % 86 44.8 % 0.00 1.00 (0.53–1.88) 0.996 IL1RN
L/L 9 31.0 % 35 36.5 % – – –
decreased risk of tumors in the non-cardia region—in codominant [OR 0.40 (0.16–0.95), p =0.046] and over-dominant [OR 0.36 (0.16–0.84), p=0.017] models. However, when taking gender into account, women car-rying theTNF-308GA genotype showed a decreased risk of non-cardia tumors [OR 0.07 (0.01–0.78), p =0.016], while in men, a decreased risk of non-cardia tumors was associated with the TNF-308GG genotype [OR 0.06 (0.01–0.47),p =0.016].
The genotypes were also analyzed in combination. Among all possible genotype combinations, only the SNPs RAD51 135G[C, XRCC3 18067C[T, and XRCC1 28152G[A showed combinations with statistically sig-nificant association with an increased risk of cardia tumors (p=0.029) (Table3). Interestingly, among the significant combinations, the presence of the RAD51 GG genotype was associated with at least three polymorphic alleles of the other two genes, while for theRAD51 CC genotype, only one polymorphic allele of the other SNPs appeared to be required to establish the relationship of risk.
Helicobacter pyloriGenotype
Almost all samples were H. pylori positive (117/127; 92.1 %). Of these, 64.9 % (76/117) werecagA?, 51.3 % (60/117) cagE?, 59 % (69/117) virB11?, 85.5 % (100/ 117)vacA s1, 70.1 % (82/117)vacA m1, 66.6 % (78/117) vacA s1m1, and 11.1 % (13/117) vacA s2m2. The distri-bution of these genes according to tumor location showed that the cagE and virB11 genes were significantly more
frequent in tumors of the non-cardia region (p =0.020 and 0.046, respectively), while other genes had the same dis-tribution in tumors from both regions. Corroborating these data, a positive correlation was found between patients infected bycagE-positiveH. pyloriand non-cardia tumors (r=0.192;p =0.030).
Discussion
An increasing incidence of adenocarcinomas of the car-dia has been observed in recent years, accompanying the increased incidence of adenocarcinomas at the esophagogastric junction mainly in the USA and Western Europe [4, 8]. Two pathways have been described for gastric cardia carcinogenesis, one associated withH. pylori atrophic gastritis, resembling non-cardia cancer, and the other associated with non-atrophic gastric mucosa, resem-bling esophageal adenocarcinoma. Also, cardia tumors are especially more frequent in males, as found in the present study [4,23].
The relationship of genetic factors with tumors located in the cardia has not been widely explored, besides the finding that the risk of gastric tumors is generally associ-ated with certain polymorphisms. In this study, considering the polymorphisms in DNA repair genes, a statistical dif-ference between cardia and non-cardia tumors was related to the SNPXRCC128152G[A, in which the heterozygous genotype (GA) was associated with a risk of the cardia region and a risk reduction for non-cardia. XRCC1 gene
Table 2 continued
Cardia Non-cardia v2 OR p TNF (-308G[A)
GG 24 80.0 % 79 81.4 %
GA 6 20.0 % 17 17.5 % 0.08 1.16 (0.36–3.61) 0.776 AA 0 0.0 % 1 1.0 % 0.30 0.00 (0.00–59.84) 1.00 G allele 54 90.0 % 175 90.2 % – – – A allele 6 10.0 % 19 9.8 % 0.00 1.02 (0.35–2.90) 0.962
IL-6 (-174G[C)
GG 9 30.0 % 16 16.7 % – – –
GC 11 36.7 % 41 42.7 % 1.94 0.48 (0.15–1.55) 0.164 CC 10 33.3 % 39 40.6 % 2.11 0.46 (0.14–1.51) 0.146 G allele 29 48.3 % 73 38.0 % – – – C allele 31 51.7 % 119 62.0 % 2.02 0.66 (0.35–1.22) 0.155
IL-8 (-251A[T)
AA 6 20.0 % 14 14.4 % – – –
encodes the key protein of the base excision repair path-way, and the SNP 28152G[A causes an amino acid change that alters the efficiency of the repair process [9,10]. One possible explanation for this association is that this geno-type (GA) leads to a less efficient repair favoring the de-velopment of adenocarcinoma in the cardia even at a low rate ofH. pyloriinfection.
The literature is contradictory about the association of XRCC128152G[A polymorphism with gastric cardia tu-mor risk. Shen et al. [11] found a borderline association with the risk of the genotypesXRCC128152 GA/AA, but the association was significant when combined with the CC genotype of the XRCC1 (26304C[T) SNP. Also, Miao et al. [12] found risk associated with XRCC1 allele A. Besides these two studies with a Chinese population showing a relative risk associated with the A allele, other studies with a Chinese population or other ethnicities found no association [13–16].
In our study, only the genotype combinations ofRAD51, XRCC3, andXRCC1 SNPswere associated with increased risk of cardia tumors. The SNP that more strongly estab-lished associations with such risk in combination with others was RAD51 135G[C. Although previous studies have not established a relationship between this SNP and cardiac gastric cancer, this polymorphism seems to be as-sociated with gastric cancer in general and esophageal squamous cell carcinoma (ESCC) [17,18]. Gastric cardia
adenocarcinoma and ESCC have many similarities (etiologic determinants and anatomical proximity), which argues for similar etiology for the two tumors [19]. RAD51 has been demonstrated to be involved in homologous re-combination repair of double-stranded DNA breaks, which may cause genomic instability and cancer [20]. Further-more, it has been demonstrated that the XRCC3–RAD51 interaction is biologically relevant, since XRCC3 protein promotes the assembly or stabilizes a multimeric form of RAD51 required for DNA repair [21].
With regard to cytokine polymorphisms, we found that the GG and GA genotypes of TNF-308G[A were less frequent in patients with tumors located in the non-cardia region. These genotypes are associated with a normal and intermediate inflammatory response, respectively [22]. Since tumors in the non-cardia region are associated with infection byH. pylori[23], the inflammation caused by this bacterium does not get worse and appears to decrease the risk of cancer development in this region. Polymorphisms of interleukins have been widely studied in gastric cancer, but few studies classify their data according to tumor location.
Data from the literature show that non-cardia gastric tumors have a strong positive association with H. pylori infection, whereas the cancer of the cardia does not have an established relationship with the presence of this microor-ganism [2]. In our study, thecagE andvirB11 genes were
Table 3 Genotype combination of SNPsRAD51,XRCC3, andXRCC1in patients with cardia and non-cardia tumors (p=0.029) RAD51 (G[C) XRCC3 (C[T) XRCC1 (G[A) Cardia/non-cardia frequencies OR
GG CC GG 0.2577/0.2333 1
GG CC GA 0.1237/0.1667 0.672 (0.1763–2.562) GG CC AA 0.06186/0.03333 1.68 (0.1724–16.37) GG CT GG 0.1443/0.06667 1.96 (0.3573–10.75) GG CT GA 0.08247/0.1667 0.448 (0.1108–1.811)
GG CT AA 0.07216/0 1.894e1010 (1.245e1010–2.88e1010) GG TT GG 0.02062/0.06667 0.28 (0.03322–2.36)
GG TT GA 0.02062/0.03333 0.56 (0.04405–7.119)
GG TT AA 0.01031/0 5.846e1007 (2.152e1007–1.588e1008) GC CC GG 0.03093/0 3.156e1009 (1.734e1009–5.744e1009) GC CC GA 0.01031/0.1 0.09333 (0.008354–1.043)
GC CC AA 0.01031/0 1.173e1008 (4.317e1007–3.186e1008) GC CT GG 0.03535/0.03333 0.96 (0.08716–10.57)
GC CT GA 0.04713/0 3.779e1010 (2.267e1010–6.302e1010) CC CC GG 0.03093/0 2.663e1009 (1.463e1009–4.847e1009)
CC CC GA 0/0.06667 0
CC CC AA 0/0.03333 0
more frequent in non-cardia tumors and cagE was positively related to them. Similar results were obtained in a previous study by our group, wherecagE-/vir B11-nega-tiveH. pyloristrains showed a slight increase in the inci-dence of gastric cardia tumors [24]. Therefore, the relationship betweenH. pyloriinfection and tumor devel-opment in non-cardia region seems to be more influenced by the virulence status of the strain than by the presence of bacteria. Thus, another carcinogenic pathway appears to act in the development of tumors of the cardia region.
In conclusion, this study indicates that there are some susceptibility differences between tumors of different lo-cations. Patients with XRCC1 28152G[A (Arg399Gln) have an increased risk of cardia tumors, also associated with some genotypes of the repair enzyme SNPs RAD51 135G[C andXRCC3 18067C[T. The G allele of the cy-tokine TNF-308G[A may negatively influence the devel-opment of non-cardia tumors. Furthermore, H. pylori strains seem to be different in the cardia region, where they are less virulent than those located in the distal region of the stomach.
Conflict of interest None.
References
1. Globocan, 2012. Available at: http://globocan.iarc.fr/. Accessed July 17, 2014.
2. Helicobacter and Cancer Collaborative Group. Gastric cancer and
Helicobacter pylori: a combined analysis of 12 case control studies nested within prospective cohorts.Gut. 2001;49:347–353. 3. Tytgat GN, Bartelink H, Bernards R, et al. Cancer of the eso-phagus and gastric cardia: recent advances. Dis Esophagus. 2004;17:10–26.
4. Piazuelo MB, Correa P. Gastric cancer: overview.Colomb Med. 2013;44:192–201.
5. Foster GD, Twell DJ. Plant gene isolation: principles and practice. England: Wiley; 1996.
6. Lage AP, Godfroid E, Fauconnier A, et al. Diagnosis of Heli-cobacter pyloriinfection by PCR: comparison with other invasive techniques and detection of cagA gene in gastric biopsy speci-mens.J Clin Microbiol. 1995;33:2752–2756.
7. Lima VP, Silva-Fernandes IJ, Alves MK, Rabenhorst SH. Prevalence ofHelicobacter pylorigenotypes (vacA, cagA, cagE and virB11) in gastric cancer in Brazilian’s patients: an asso-ciation with histopathological parameters. Cancer Epidemiol. 2011;35:32–37.
8. de Martel C, Ferlay J, Franceschi S, et al. The global burden of cancers attributable to infections in the year 2008: a review and synthetic analysis.Lancet Oncol. 2012;13:607–615.
9. Lunn RM, Langlois RG, Hsieh LL, Thompson CL, Bell DA. XRCC1 polymorphisms: effects on aflatoxin B1-DNA adducts and gly-cophorin A variant frequency.Cancer Res. 1999;59:2557–2561. 10. Duell EJ, Wiencke JK, Cheng TJ, et al. Polymorphisms in the
DNA repair genes XRCC1 and ERCC2 and biomarkers of DNA damage in human blood mononuclear cells. Carcinogenesis. 2000;21:965–971.
11. Shen H, Xu Y, Qian Y, et al. Polymorphisms of the DNA repair gene XRCC1 and risk of gastric cancer in a Chinese population.
Int J Cancer. 2000;88:601–606.
12. Miao X, Zhang X, Zhang L, et al. Adenosine diphosphate ribosyl transferase and x-ray repair cross-complementing 1 polymorphisms in gastric cardia cancer.Gastroenterology. 2006;131:420–427. 13. Ratnasinghe LD, Abnet C, Qiao YL, et al. Polymorphisms of
XRCC1 and risk of esophageal and gastric cardia cancer.Cancer Lett. 2004;216:157–164.
14. Yan L, Yanan D, Donglan S, Na W, Rongmiao Z, Zhifeng C. Polymorphisms of XRCC1 gene and risk of gastric cardiac ade-nocarcinoma.Dis Esophagus. 2009;22:396–401.
15. Xue H, Ni P, Lin B, Xu H, Huang G. X-ray repair cross-com-plementing group 1 (XRCC1) genetic polymorphisms and gastric cancer risk: a HuGE review and meta-analysis.Am J Epidemiol. 2011;173:363–375.
16. Chen B, Zhou Y, Yang P, Wu XT. Polymorphisms of XRCC1 and gastric cancer susceptibility: a meta-analysis.Mol Biol Rep. 2012;39:1305–1313.
17. Poplawski T, Arabski M, Kozirowskaa D, et al. DNA damage and repair in gastric cancer—a correlation with the hOGG1 and RAD51 genes polymorphisms.Mutat Res. 2006;601:83–91. 18. Fan XJ, Ren PL, Lu ZJ, Zhao S, Yang XL, Liu J. The study of
esophageal cancer risk associated with polymorphisms of DNA damage repair genes XRCC4 and RAD51.Sichuan Da Xue Xue Bao Yi Xue Ban. 2013;44:568–572.
19. Tramacere I, Pelucchi C, Bagnardi V, et al. A meta-analysis on alcohol drinking and esophageal and gastric cardia adenocarci-noma risk.Ann Oncol. 2012;23:287–297.
20. Shrivastav M, De Haro LP, Nickoloff JA. Regulation of DNA double-strand break repair pathway choice. Cell Res. 2008;18: 134–147.
21. Bishop DK, Ear U, Bhattacharyya A, et al. XRCC3 is required for assembly of RAD51 complexes in vivo.J Biol Chem. 1998;21: 21482–21488.
22. Wilson AG, Symons JA, McDowell TL, McDevitt HO, Duff GW. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation.Proc Natl Acad Sci USA. 1997;94:3195–3199.
23. Hansen S, Vollset SE, Derakhshan MH, et al. Two distinct aeti-ologies of cardia cancer; evidence from premorbid serological markers of gastric atrophy and Helicobacter pyloristatus.Gut. 2007;56:918–925.
